DESCRIPTION: (Applicant's abstract): Most current models of word recognition emphasize the role of semantic information in the representation and processing of lexical items. Moreover, most of these models are relatively vague in specifying computational processes and features, particularly with regard to non-semantic cues such as phonological, orthographic and prosodic information. Two models, however, more explicitly address the use of non-semantic information: the autonomous search model proposes the action of three peripheral access files comprising phonological, orthographic and semantic/syntactic information, which refer to lexical information in a "master" file. TRACE theory describes the process of an item in terms of excitatory and inhibitory interactions of nodes at different hierarchical levels of information. The types of non-semantic information used in such models has not been clearly specified, due to the lack of relevant evidence. To fill in these empirical and theoretical gaps, the present research will use a priming paradigm to examine three types of non-semantic cues (phonological, orthographic and prosodic) in the processing of auditory and visually presented words. These three cues will be systematically examined using lexical decision, naming, word identification, phoneme monitoring and word monitoring tasks. The research is organized into three stages that will examine 1) the use of each individual cue in word recognition; 2) the extent to which these cues interact; and, 3) the effect of these cues on word recognition in context. By conducting a series of closely related experiments, data will be obtained which reveal the nature of these cues in the recognition process. As a result, the impact of these cues relative to the better understood semantic cues will be gauged. Moreover, an understanding of these cues and their inclusion in models of word recognition may ultimately lead to the application of techniques emphasizing appropriate cues in educational and language programs.